the influence of mesh density on the impact response of a

November 2004

NASA/TM-2004-213501

ARL-TR-3337

The Influence of Mesh Density on the Impact

Response of a Shuttle Leading-Edge Panel

Finite Element Simulation

Karen E. Jackson and Edwin L. Fasanella

U.S. Army Research Laboratory

Vehicle Technology Directorate

Langley Research Center, Hampton, Virginia

Karen H. Lyle and Regina L. Spellman


The NASA STI Program Office . . . in Profile

Since its founding, NASA has been dedicated to the

advancement of aeronautics and space science. The

NASA Scientific and Technical Information (STI)

Program Office plays a key part in helping NASA

maintain this important role.

The NASA STI Program Office is operated by

Langley Research Center, the lead center for NASA’s

scientific and technical information. The NASA STI

Program Office provides access to the NASA STI

Database, the largest collection of aeronautical and

space science STI in the world. The Program Office is

also NASA’s institutional mechanism for

disseminating the results of its research and

development activities. These results are published by

NASA in the NASA STI Report Series, which

includes the following report types:

• TECHNICAL PUBLICATION. Reports of

completed research or a major significant phase

of research that present the results of NASA

programs and include extensive data or

theoretical analysis. Includes compilations of

significant scientific and technical data and

information deemed to be of continuing

reference value. NASA counterpart of peer-

reviewed formal professional papers, but having

less stringent limitations on manuscript length

and extent of graphic presentations.

• TECHNICAL MEMORANDUM. Scientific

and technical findings that are preliminary or of

specialized interest, e.g., quick release reports,

working papers, and bibliographies that contain

minimal annotation. Does not contain extensive

analysis.

• CONTRACTOR REPORT. Scientific and

technical findings by NASA-sponsored

contractors and grantees.

• CONFERENCE PUBLICATION. Collected

papers from scientific and technical

conferences, symposia, seminars, or other

meetings sponsored or co-sponsored by NASA.

• SPECIAL PUBLICATION. Scientific,

technical, or historical information from NASA

programs, projects, and missions, often

concerned with subjects having substantial

public interest.

• TECHNICAL TRANSLATION. English-

language translations of foreign scientific and

technical material pertinent to NASA’s mission.

Specialized services that complement the STI

Program Office’s diverse offerings include creating

custom thesauri, building customized databases,

organizing and publishing research results ... even

providing videos.

For more information about the NASA STI Program

Office, see the following:

• Access the NASA STI Program Home Page at

http://www.sti.nasa.gov

• E-mail your question via the Internet to

[email protected]

• Fax your question to the NASA STI Help Desk

at (301) 621-0134

• Phone the NASA STI Help Desk at

(301) 621-0390

• Write to:

NASA STI Help Desk

NASA Center for AeroSpace Information

7121 Standard Drive

Hanover, MD 21076-1320

National Aeronautics and

Space Administration

Langley Research Center

Hampton, Virginia 23681-2199

November 2004

NASA/TM-2004-213501

ARL-TR-3337

The Influence of Mesh Density on the Impact

Response of a Shuttle Leading-Edge Panel

Finite Element Simulation


U.S. Army Research Laboratory

Vehicle Technology Directorate




Available from:

NASA Center for AeroSpace Information (CASI) National Technical Information Service (NTIS)

7121 Standard Drive 5285 Port Royal Road

Hanover, MD 21076-1320 Springfield, VA 22161-2171

(301) 621-0390 (703) 605-6000

1

The Influence of Mesh Density on the Impact Response of a Shuttle

Leading-Edge Panel Finite Element Simulation


US Army Research Laboratory, VTD

Hampton, VA


NASA Langley Research Center

Hampton, VA

Abstract

A study was performed to examine the influence of varying mesh density on an LS-

DYNA simulation of a rectangular-shaped foam projectile impacting the space shuttle

leading edge Panel 6. The shuttle leading-edge panels are fabricated of reinforced

carbon-carbon (RCC) material. During the study, nine cases were executed with all

possible combinations of coarse, baseline, and fine meshes of the foam and panel. For

each simulation, the same material properties and impact conditions were specified and

only the mesh density was varied. In the baseline model, the shell elements representing

the RCC panel are approximately 0.2-in. on edge, whereas the foam elements are about

0.5-in. on edge. The element nominal edge-length for the baseline panel was halved to

create a fine panel (0.1-in. edge length) mesh and doubled to create a coarse panel (0.4-

in. edge length) mesh. In addition, the element nominal edge-length of the baseline foam

projectile was halved (0.25-in. edge length) to create a fine foam mesh and doubled (1.0-

in. edge length) to create a coarse foam mesh. The initial impact velocity of the foam

was 775 ft/s. The simulations were executed in LS-DYNA version 960 for 6 ms of

simulation time. Contour plots of resultant panel displacement and effective stress in the

foam were compared at five discrete time intervals. Also, time-history responses of

internal and kinetic energy of the panel, kinetic and hourglass energy of the foam, and

resultant contact force were plotted to determine the influence of mesh density. As a

final comparison, the model with a fine panel and fine foam mesh was executed with

slightly different material properties for the RCC. For this model, the average degraded

properties of the RCC were replaced with the maximum degraded properties. Similar

comparisons of panel and foam responses were made for the average and maximum

degraded models.

Introduction

Following the Space Shuttle Columbia disaster on February 1, 2003 and during the

subsequent investigation by the Columbia Accident Investigation Board (CAIB), various

teams from industry, academia, national laboratories, and NASA were requested by

Johnson Space Center (JSC) Orbiter Engineering to apply “physics-based” analyses to

characterize the expected damage to the shuttle thermal protection system (TPS) tile and

Reinforced Carbon-Carbon (RCC) material, for high-speed foam impacts. The forensic

2

evidence from the Columbia debris eventually led investigators to conclude that the

breach to the shuttle TPS was caused by a large piece of External Tank (ET) foam that

impacted and penetrated the lower portion of a left-wing leading-edge panel, shown in

Figure 1. As a result, NASA authorized a series of tests that were performed at

Southwest Research Institute to characterize the impact response of the leading-edge

RCC panels.

Recommendation 3.3-2 of the CAIB report [1] requests that NASA initiate a program to

improve the impact resistance of the wing leading edge. The second part of the

recommendation is to …“determine the actual impact resistance of current materials and

the effect of likely debris strikes.” For Return-to-Flight (RTF), a team consisting of

personnel from NASA Glenn Research Center, NASA Langley Research Center, and

Boeing Philadelphia was given the following task: to develop a validated finite element

model of the shuttle wing leading edge capable of accurately predicting the threshold of

damage from debris including foam, ice, and ablators for a variety of impact conditions.

Since the CAIB report was released, the team has been developing finite element models

of the RCC leading-edge panels; executing the models using LS-DYNA [2], a

commercial nonlinear explicit transient dynamic finite element code; conducting detailed

material characterization tests to obtain dynamic material property data; and, correlating

the LS-DYNA analytical results with experimental data obtained from impacts tests onto

RCC panels. Some of the early results of this research are described in References 3-7.

22 panels per wing

Panel 6

Figure 1. Drawing of the left wing area of the space shuttle.

The purpose of this report is to describe a mesh density study that was performed as part

of the ongoing RTF modeling efforts. In particular, the mesh study was focused on

simulating a rectangular foam projectile, having the same material properties as the BX-

250 foam used on the shuttle ET, impacting the shuttle leading-edge RCC Panel 6 at a

velocity of 775 ft/s. The location of Panel 6 on the left wing of the shuttle is highlighted

3

in Figure 1. An actual impact test of a BX-250 foam block onto RCC Panel 6 was

performed at Southwest Research Institute on June 5, 2003. While the finite element

model was originally developed to generate analytical predictions for correlation with

experimental data obtained from this test, the focus of the mesh density study described

in this report is strictly analytical.

For the mesh discretization study, nine cases were executed with all possiblecombinations of coarse, baseline, and fine meshes of the foam and panel. For eachsimulation, the same material properties and impact conditions were specified and onlythe mesh density was varied. The simulations were executed in LS-DYNA version 960for 6 ms of simulation time. Predicted structural deformations and time-history responsesare compared for each simulation. As a final comparison, the model with a fine paneland fine foam mesh was executed with slightly different material properties for the RCC.For this model, the average degraded properties of the RCC were replaced with themaximum degraded properties. Comparisons of panel deformation, effective stress in thefoam, and selected time-history responses were made for the average and maximumdegraded models.

Model Description

The complete model including the foam projectile and the RCC Panel 6 is shown in

Figure 2. The Panel 6 model was discretized using Belytachko-Tsay quadrilateral shell

elements having nominal element edge lengths of 0.1-, 0.2-, and 0.4-inches for the fine,

baseline, and coarse meshes, respectively. A schematic illustrating the different mesh

densities for the panel is shown in Figure 3. The panel model consisted of 19 different

parts including the panel midsection, two bottom flanges, two side ribs, two apex ribs,

and twelve bolt-holes. These parts are labeled in Figure 4.

Figure 2. Foam projectile and shuttle RCC Panel 6 model.

4

1.6-in.

1.6-in.

Coarse panel mesh

Fine panel mesh1.6-in.

Baseline panel mesh

Figure 3. Comparison of Panel 6 mesh densities.

Apex 1 & 2

Midsection

Ribs 1 & 2Bottom flange Bolt-hole

constraints, typical

Figure 4. Part designations for the RCC Panel 6 model.

The quadrilateral shell elements representing the RCC panel midsection and ribs were

assigned material type 58, designated MAT_LAMINATED_COMPOSITE_FABRIC.

These parts were modeled as a 19-ply laminated composite fabric with the fibers in each

layer oriented in the 0°/90° direction. The bottom flanges were modeled as a 25-ply

laminated composite fabric, having slightly different stiffness and strength properties for

the RCC material. The rib apex parts were also modeled as a 19-ply laminated composite

fabric again having slightly different properties than the RCC material assigned to the

midsection and ribs. Three unique material designations were used to specify the

material properties of the RCC in the model. These three specifications were needed to

account for differences in flight conditioned, mass degraded, and damaged material

states.

5

For each RCC material designation, average degraded material properties were used.

Prior testing of RCC material shows that it is much stiffer and stronger in compression

than in tension, thus requiring a bimodular material model. Also, the stiffness and

strength of pristine RCC material are significantly higher than flight-conditioned

material. Consequently, the term ‘degraded’ refers to the fact that flight-conditioned

material properties were used. RCC also exhibits considerable variability in material

response and it is common to see a band or range of curves used to describe the tensile

and/or compressive response, typically maximum, average, and minimum response

curves. For this study, the term ‘average’ means that the average curve was chosen for

input. In the final analysis performed in this study, a comparison is made between

average and maximum degraded properties. Thus, the term ‘maximum’ refers to the

curve defining the upper limit of the range.

In the actual RCC Panel 6, bolts were used to support and constrain the panel at the bolt-

hole locations. To account for the constraint provided by the bolts in the model, the bolt-

holes were represented using 0.1-in.-thick shell elements that were assigned rigid

material properties using material type 20 MAT_RIGID. Then, these elements were

constrained from translational motion in the x-, y-, and z-directions using the

BOUNDARY_PRESCRIBED_MOTION_RIGID card in LS-DYNA.

The finite element model of the BX-250 foam projectile had overall dimensions of 5.5 x

11.5 x 22.5-in. and was discretized using hexagonal solid elements having nominal

element edge lengths of 0.25-, 0.5-, and 1.0-in. for the fine, baseline, and coarse meshes,

respectively. A schematic illustrating the different foam mesh densities is shown in

Figure 5. Also, a comparison of the total number of nodes and elements in the three foam

and three panel meshes is provided in Table 1. The foam block represented a single part

in the LS-DYNA model, making the total number of parts in the model equal to 20. The

foam block weighed 1.67 lb.

The material properties of the BX-250 foam were represented using material type 83

MAT_FU_CHANG_FOAM with MAT_ADD_EROSION in LS-DYNA. The erosion

card is added to allow for element failure in the foam constitutive model. The

experimental foam material responses were input into the model using the

DEFINE_CURVE command in LS-DYNA. The responses were obtained from the

testing of foam components performed at NASA Glenn Research Center. These tests

were conducted to determine the influence of strain rate on the compressive response of

the foam material. Results for two strain rates, 0.01 s-1

and 25 s-1

, are plotted in Figure 6.

The material response data are plotted only up to 200-psi stress to aid in visualization of

the differences caused by strain rate; however, the stress data at strain values approaching

1 are 70,000 psi and higher. The response of the BX-250 foam, shown in Figure 6, is

typical of other foam responses in that it exhibits a linear response at low strains, and as

crushing begins a “knee” occurs in the response. Then, as stable crushing continues, the

stress increases gradually until the cells within the foam begin to compact. As

compaction initiates and continues, the stress increases dramatically for relatively small

increases in strain. As shown in Figure 6, the influence of strain rate is to increase the

stress at which the knee occurs, to increase the stress during stable crushing, and to lower

6

the strain at which compaction begins. A tensile failure stress of 65-psi was assigned to

the foam.

Coarse foam mesh

1.0-in.

1.0-in.

Fine foam mesh1.0-in.

Baseline foam mesh

Figure 5. Comparison of mesh densities used for the foam.

Table 1. Comparison of the number of elements and nodes per panel and foam mesh.

Panel Foam

Mesh density Number of

elements

Number

of nodes

Number of

elements

Number of

nodes

Coarse 11,170 11,459 1,380 1,872

Baseline 32,109 32,432 11,385 13,248

Fine 128,172 141,723 92,092 99,452

0

50

100

150

200

0 0.2 0.4 0.6 0.8 1

0.01 s-1

strain rate

25.0 s-1

strain rate

Stress, psi

Strain, in/in

Figure 6. Compressive material properties of BX-250 foam for two different strain rates.

7

All of the nodes used to create the foam projectile were assigned an initial velocity of 775

ft/s (9,300 in/s) in the x-direction, which is defined along the long edge of the foam block

(see Figure 2). A CONTACT_ERODING_NODES_TO_SURFACE was specified

between the panel midsection and the foam in the model. For this contact, the panel

midsection was designated the master surface, and the foam was the slave. Due to the

eroding feature of this contact definition, a foam element may fail, or erode, and the

contact will be picked up by the next element.

For this analytical study, nine simulations were performed for every combination of

coarse, baseline, and fine mesh of the foam and RCC Panel 6. The matrix of simulations

and the naming scheme used to differentiate the models are shown in Figure 7. For these

nine simulations, the same material properties were used for the foam and panel and the

same initial velocity and contact definitions were specified. Only the mesh densities

were varied. Each model was executed for 0.006 s (6 ms) of simulation time using LS-

DYNA Version 960. The simulations were run on a single-processor Linux-based

Hewlett Packard workstation x4000.

Coarse Panel:Coarse Foam(CP:CF)

Fine Panel:Coarse Foam(FP:CF)

Baseline Panel:Baseline Foam(BP:BF)

Fine Panel:Baseline Foam(FP:BF)

Coarse Panel:Baseline Foam(CP:BF)

Baseline Panel:Coarse Foam(BP:CF)

Baseline Panel:Fine Foam(BP:FF)

Fine Panel:Fine Foam(FP:FF)

Coarse Panel:Fine Foam(CP:FF)

Figure 7. Analysis matrix.

Simulation Results

The results of the analytical study are presented as contour plots of resultant panel

deflections, contour plots of effective stress in the foam, and time-history plots of internal

and kinetic energy of the panel midsection, the resultant contact force response, and the

kinetic and hourglass energy of the foam.

Contour Plots of Predicted Resultant Panel Deflections

The predicted resultant panel deflections are shown in Figures 8 through 12 at discrete

time intervals of 1.8, 2.8, 3.8, 5.0, and 6.0 ms, respectively. The results are shown using

the format illustrated in Figure 7, and the maximum deflection values are provided in the

parenthesis beneath each contour plot at each time interval. Note that the contour plots

8

are shown for the same fringe levels at each time interval; however, the maximum value

of the range is different for each time interval. To orient the reader, the contour plots

shown in Figures 8-12 were created by turning the panel on its edge and viewing the

lower surface of the panel midsection through the bottom flanges.

At 1.8- and 2.8-ms, the trend is that, for a constant panel mesh, increases in foam mesh

density cause a decrease in maximum deflection. However, for a constant foam mesh,

increases in panel mesh density produce an increase in maximum deflection. Given this

trend, the fine panel:coarse foam (FP:CF) model exhibits the maximum deflection of all

the combinations of foam and panel meshes for these two time intervals. No failure of

the panel has occurred by 2.8 ms, as shown in Figure 9.

CP:CF (0.24-in.) CP:BF (0.23-in.) CP:FF (0.22-in.)

BP:CF (0.25-in.) BP:BF (0.23-in.) BP:FF (0.23-in.)

FP:CF (0.27-in.) FP:BF (0.25-in.) FP:FF (0.24-in.)

Figure 8. Resultant deflection plots of RCC Panel 6 at 1.8 ms.

9





By 3.8 ms, all of the finely meshed panels have failed, regardless of foam mesh density.

For the fine panel with coarse and baseline foam meshes, failure initiates as a crack in the

rib/apex intersection region, as observed in Figure 10. However, failure of the fine panel

with fine foam (FP:FF) initiates as a crack in the panel midsection. For this model, the

rib/apex crack develops as a secondary damage mode. In general, the FP:FF model

exhibits more extensive damage at this time than seen in the other models. Note that

none of the coarse or baseline panels have failed at this time.

10





By 5.0 ms, all of the panels have failed with cracking of the rib area as a common

damage mode, as seen in Figure 11. In all but two cases, the rib cracks initiate at the

interface region between the rib and apex, where the material properties change. The two

exceptions are the baseline panel:baseline foam (BP:BF) and the baseline panel:fine foam

(BP:FF). For these two models, the crack initiates in the rib area only, away from the

rib/apex interface. The rib crack is the only damage seen in the coarse and baseline

panels at this time step. In comparison, the rib crack in the fine panel models has grown

downward, separating a portion of the rib from the panel. By 5.0 ms, the panel

midsection has failed in all of the fine panel models. The panel failure is characterized

by the formation of a large crack in the panel midsection that runs parallel to the rib/panel

interface. Finally, the FP:FF model exhibits a second crack in the panel midsection that

runs normal to the rib/panel interface. A similar crack is not observed in the fine panel

models with coarse or baseline foam meshes.

11

CP:CF (1.12-in.)CP:BF (1.06-in.) CP:FF (1.05-in.)




By 6.0 ms, the rib/apex cracks in all of the coarse panel models and in the BP:CF model

have grown downward along the rib/panel interface, as shown in Figure 12. The rib

cracks in the baseline panel models with baseline and fine foam are stable, exhibiting no

increase in the crack size or formation of new damage. The damage in the fine panel

models is also stabilized.

12





Contour Plots of Predicted Effective Stress in the Foam ProjectileContour plots of effective stress for varying foam densities are shown in Figures 13-15,

for constant coarse, baseline, and fine panel meshes, respectively. Note that the same

fringe levels are used in each figure. The plots show that the amount of damage in the

foam increases as the mesh density of the foam increases, for a constant panel mesh.

Thus, the fine foam model exhibits the maximum damage in each case. This trend is

observed for all panel meshes.

Contour plots of effective stress for a constant baseline foam mesh are shown in Figure

16 for varying panel densities. As illustrated in the figure, the stress levels in the baseline

foam at each time interval are nearly identical, regardless of panel mesh. However, the

CP:BF model exhibits more damage than the BP:BF or FP:BF models at 5 and 6 ms.

13

Time,

ms

Coarse Foam Baseline Foam Fine Foam Fringe

Info

1.8

2.8

3.8

5.0

6.0

Figure 13. Fringe plots of effective stress for varying foam densities given a constant

panel mesh (coarse).

14

Time,

ms


Info

1.8

2.8

3.8

5.0

6.0

Figure 14. Contour plots of effective stress for varying foam mesh densities given a

constant panel mesh (baseline).

15

Time,

ms


Info

1.8

2.8

3.8

5.0

6.0

Figure 15. Contour plots of effective stress for varying foam mesh densities given a

constant panel mesh (fine).

16

Time,

ms

Coarse Panel Baseline Panel Fine Panel Fringe

Info

1.8

2.8

3.8

5.0

6.0

Figure 16. Contour plots of effective stress for a constant baseline foam mesh and

varying panel mesh densities.

Time-History Response Comparisons

Comparisons of predicted time-history responses for internal and kinetic energy of the

panel, kinetic and hourglass energy of the foam, and resultant contact force are shown in

Figures 17-19 for coarse, baseline, and fine panel meshes with varying foam meshes,

respectively.

The time-history results for the coarse panel with varying foam meshes are shown in

Figure 17. The internal and kinetic energy responses of the panel and the kinetic energy

responses of the foam show only minor differences due to varying foam mesh densities.

No differences in the overall magnitude and duration of the resultant contact force time-

history are seen, as shown in Figure 17 (d); however, the coarse foam response contains

high-frequency oscillations, while the baseline and fine foam responses are smooth. As

seen in Figure 17 (e), the coarse foam response exhibits significantly higher hourglass

energy that the baseline or fine foam responses. Generally, this finding would make the

coarse foam mesh undesirable; however, the magnitude of the hourglass energy is small

when compared with the kinetic energy of the foam.

17

-200

0

200

400

600

800

1000

1200

1400

0 0.001 0.002 0.003 0.004 0.005 0.006

Coarse foamBaseline foamFine foam

Inte

rnal

Ene

rgy,

in-lb

.

Time, s

-100

0

100

200

300

400

0 0.001 0.002 0.003 0.004 0.005 0.006

Baseline foamCoarse foam

Fine foam

Kin

etic

Ene

rgy,

in-lb

.

Time, s

(a) Internal energy of the panel. (b) Kinetic energy of the panel.

165000

170000

175000

180000

185000

190000

0 0.001 0.002 0.003 0.004 0.005 0.006


Kin

etic

Ene

rgy,

in-lb

.

Time, s

(c) Kinetic energy of the foam.

-1000

0

1000

2000

3000

4000

0 0.001 0.002 0.003 0.004 0.005 0.006

Baseline foamCoarse foam

Fine foam

Con

tact

For

ce, l

b.

Time, s

-1000

0

1000

2000

3000

4000

5000

0 0.001 0.002 0.003 0.004 0.005 0.006


Hou

rgla

ss E

nerg

y, in

-lb.

Time, s

(d) Contact force. (e) Hourglass energy of the foam.

Figure 17. Predicted time-history responses for a constant coarse panel mesh and varying

foam densities.

18

Similar trends are observed in the time-history responses of the baseline panel with

varying foam mesh densities, plotted in Figure 18. The primary difference between the

coarse and baseline panel responses is that the hourglass energy of the coarse foam curve

is lower for the baseline case. However, the amount of hourglass energy in the coarse

foam is still greater than that in the baseline or fine foam.

The time-history responses for the fine panel with varying foam mesh densities are

shown in Figure 19. Unlike the coarse and baseline panels where only minor differences

were seen in the internal and kinetic energy responses of the panel as a function of

varying foam mesh density, the fine panel results show much larger variations in these

responses, as shown in Figures 19 (a) and (b). In both cases, the FP:CF model exhibits

the highest magnitude response. However, only minor differences in the kinetic energy

responses of the foam, shown in Figure 19 (c), are observed for the differing foam mesh

densities. The resultant contact force responses are of equal magnitude and duration, see

Figure 19 (d), and the coarse foam response exhibits high-frequency oscillations, not seen

in the other responses. The large spike in the fine foam response is attributed to

instabilities in the contact algorithm. Finally, the hourglass energy responses are shown

in Figure 19 (e), with the coarse foam response exhibiting the highest amount of

hourglass energy. However, even for the coarse mesh, the magnitude of the hourglass

energy is only a fraction of the kinetic energy of the foam.

The final time-history comparisons are shown in Figure 20 for a constant baseline foam

mesh with a varying panel mesh. All of the responses are remarkably similar in

magnitude and shape, except for the kinetic energy time-histories of the panel, shown in

Figure 20 (b). In this case, the shapes of the curves vary considerably after 0.0015 s, with

the fine panel model exhibiting the highest magnitude response.

Influence of RCC Material Properties

As a final investigation, a simulation was executed of the fine panel:fine foam (FP:FF)

model in which the maximum degraded properties were used to represent the RCC

material. Comparisons of the FP:FF model executed using the average versus the

maximum degraded RCC properties are shown in Figures 21-23. The contour plots of

resultant panel displacement are shown in Figure 21 for the FP:FF model with average

and maximum degraded RCC properties. Below each plot, the maximum deflection at

that time step is shown in the parenthesis. For every time step, the maximum value of

resultant displacement is higher for the average degraded property case. Also, the

contour plots show that the overall amount of damage is significantly less for the

maximum degraded model than for the average degraded model.

19

-200

0

200

400

600

800

1000

1200

1400

0 0.001 0.002 0.003 0.004 0.005 0.006


Inte

rnal

Ene

rgy,

in-lb

.

Time, s

-100

0

100

200

300

400

500

0 0.001 0.002 0.003 0.004 0.005 0.006


Kin

etic

Ene

rgy,

in-lb

.

Time, s


165000

170000

175000

180000

185000

190000

0 0.001 0.002 0.003 0.004 0.005 0.006


Kin

etic

Ene

rgy,

in-lb

.

Time, s


-1000

0

1000

2000

3000

4000

0 0.001 0.002 0.003 0.004 0.005 0.006


Con

tact

For

ce, l

b.

Time, s

-1000

0

1000

2000

3000

4000

5000

0 0.001 0.002 0.003 0.004 0.005 0.006


Hou

rgla

ss E

nerg

y, in

-lb.

Time, s


Figure 18. Predicted time-history responses for a constant baseline panel mesh and

varying foam densities.

20

-200

0

200

400

600

800

1000

1200

1400

0 .001 .002 .003 .004 .005 .006

Coarse foam

Fine FoamBaseline Foam

Inte

rnal

Ene

rgy,

in-lb

.

Time, s

-100

0

100

200

300

400

500

0 0.001 0.002 0.003 0.004 0.005 0.006

Coarse foamBaseline foamFine foamK

inet

ic E

nerg

y, in

-lb.

Time, s


170000

175000

180000

185000

190000

0 0.001 0.002 0.003 0.004 0.005 0.006


Kin

etic

Ene

rgy,

in-lb

.

Time, s


-1000

0

1000

2000

3000

4000

0 0.001 0.002 0.003 0.004 0.005 0.006


Con

tact

For

ce, l

b.

Time, s

-1000

0

1000

2000

3000

4000

5000

0 0.001 0.002 0.003 0.004 0.005 0.006


Hou

rgla

ss E

nerg

y, in

-lb.

Time, s


Figure 19. Predicted time-history responses with a constant fine panel mesh and varying

foam densities.

21

-200

0

200

400

600

800

1000

1200

1400

0 0.001 0.002 0.003 0.004 0.005 0.006

Coarse panelBaseline panelFine panel

Inte

rnal

Ene

rgy,

in.-

lb.

Time, s

-100

0

100

200

300

400

500

0 0.001 0.002 0.003 0.004 0.005 0.006


Kin

etic

Ene

rgy,

in-lb

.

Time, s


170000

175000

180000

185000

190000

0 0.001 0.002 0.003 0.004 0.005 0.006


Kin

etic

Ene

rgy,

in-lb

.

Time, s


-500

0

500

1000

1500

2000

2500

3000

3500

0 0.001 0.002 0.003 0.004 0.005 0.006


Con

tact

For

ce, l

b.

Time, s

-500

0

500

1000

1500

2000

2500

3000

3500

0 0.001 0.002 0.003 0.004 0.005 0.006


Hou

rgla

ss E

nerg

y, in

-lb.

Time, s


Figure 20. Predicted time-history responses with a constant baseline foam mesh and

varying panel densities.

22

Time,ms

Average DegradedRCC

MaximumDegraded RCC

FringeLevels

1.8

FP:FF (0.24-in.) FP:FF (0.23-in.)

2.8

FP:FF (0.93-in.) FP:FF (0.75-in.)

3.8

FP:FF (1.96-in.) FP:FF (1.43-in.)

5.0

FP:FF (3.05-in.)FP:FF (1.62-in.)

6.0

FP:FF (2.89-in.) FP:FF (0.91-in.)

Figure 21. Contour plots of resultant displacement for the FP:FF model.

23

The contour plots of effective stress in the foam are compared for the maximum and

average degraded models in Figure 22 for five discrete time steps. The distribution of

effective stress and the overall amount of damage in the foam are similar in both models

at each time step.

Time,

ms

Average Degraded

RCC

Maximum

Degraded RCC

Fringe

Info

1.8

2.8

3.8

5.0

6.0

Figure 22. Fringe plots of effective stress for fine foam for average and maximum

degraded RCC material properties.

Finally, the time-history plots of internal and kinetic energy of the panel, resultant contact

force, and kinetic and hourglass energy of the foam are shown in Figure 23 for both the

maximum and average degraded models. The internal energy curves of the panel, shown

in Figure 23 (a), are closely matched for both models up to 0.004 s, after which time the

two curves diverge. The kinetic energy responses of the panel, shown in Figure 23 (b),

are close up to 0.002 s, after which time, the kinetic energy of the average degraded

model continues to increase and peaks at a maximum value of 400-in-lb. and then

decreases, whereas the maximum degraded curve peaks initially at 250-in-lb., falls off,

and then increases again to a peak of 300-in-lb by 0.006 s. The kinetic energy responses

of the foam, plotted in Figure 23 (c), are nearly identical for both the average and

maximum degraded material models. Likewise, the contact force and hourglass energy

24

responses for the two models are very similar, as shown in Figures 23 (d) and (e). Note

that both time-history responses exhibit a large spike in the contact force due to

instabilities in the contact algorithm.

-200

0

200

400

600

800

1000

0 0.001 0.002 0.003 0.004 0.005 0.006

Avg. degraded propertiesMax. degraded propertiesIn

tern

al E

nerg

y, in

-lb.

Time, s

-100

0

100

200

300

400

500

0 0.001 0.002 0.003 0.004 0.005 0.006

Avg. degraded propertiesMax. degraded propertiesK

inet

ic E

nerg

y, in

-lb.

Time, s


170000

175000

180000

185000

190000

0 0.001 0.002 0.003 0.004 0.005 0.006

Avg. degraded propertiesMax. degraded properties

Kin

etic

Ene

rgy,

in-lb

.

Time, s


-500

0

500

1000

1500

2000

2500

3000

3500

0 0.001 0.002 0.003 0.004 0.005 0.006

Avg. degraded propertiesMax. degraded properties

Con

tact

For

ce, l

b.

Time, s

-500

0

500

1000

1500

2000

2500

0 0.001 0.002 0.003 0.004 0.005 0.006

Avg. degraded propertiesMax. degraded propertiesH

ourg

lass

Ene

rgy,

in-lb

.

Time, s


Figure 23. Comparison of predicted time-history responses for the FP:FF model with

average versus degraded RCC Material properties.

25

Discussion of Results

An important factor for consideration in this study is the computational expense required

to run the fine foam and fine panel models. The time step used in executing the model is

equal to the time it takes a sound wave to cross some characteristic length, which is a

function of the smallest element in the model. Thus, the finer the mesh, the smaller the

time step. For example, the initial time step for the CP:CF model was 2.9E-7 s and the

time step for the FP:FF model was 1.4E-7 s. These two models required 4 and 101 hours

of CPU for execution, respectively. For comparison, the BP:BF model had an initial time

step of 2.7E-7 s and required 14 hours of CPU. The baseline mesh discretization is a

good choice for future simulations in that it captures the structural behavior and damage

progression of the panel and foam without the computational expense of the fine mesh

discretization. As shown in Figures 17-19, the coarse foam mesh should be avoided due

to the high levels of hourglass energy in the model.

Another important finding of this study is highlighted in Figure 21, which shows a

comparison of contour plots of resultant panel displacement for two FP:FP models, one

with average degraded properties for the RCC material and the other with maximum

degraded properties. This comparison shows that input of accurate material properties

for the RCC is absolutely necessary to correctly predict the initiation and progression of

impact damage in the panel.

Concluding Remarks

A mesh density study was performed in support of the Shuttle Return-To-Flight program

based on an LS-DYNA simulation of a foam projectile impacting one of the reinforced

carbon-carbon (RCC) leading edge panels on the space shuttle (Panel 6). For the study,

three meshes (coarse, baseline, and fine) were used for the foam and the panel. Thus,

nine simulations were executed representing all possible combinations of foam and panel

meshes. For each simulation, the same material properties and impact conditions werespecified and only the mesh density was varied. Comparisons of contour plots of resultantpanel displacement and effective stress in the foam projectile were made for five discretetime intervals. Also, time-history responses of internal and kinetic energy of the panel,kinetic and hourglass energy of the foam, and the resultant contact force were plotted todetermine the influence of mesh density. As a final comparison, the model with a finepanel and fine foam mesh was executed with slightly different material properties for theRCC material. For this model, the average degraded material properties of the RCC werereplaced with the maximum degraded properties. Comparisons of similar analyticalresults were made between the average and maximum degraded models.

The findings from this study are listed, as follows:

1. Before failure, the trend is that increasing panel mesh density results in increasingmaximum deflection for a constant foam discretization.

2. However, the opposite is true for increasing foam mesh density, which results inlower maximum deflections for a constant panel discretization.

26

3. For any foam mesh, the fine panel fails by 3.8 ms, with the FP:FF modelexhibiting the largest amount of damage, i.e. panel failure followed by ribcracking.

4. By 5.0 ms, all panels have failed, with rib cracking as the common failure mode.5. In all but two cases, the rib cracks initiate at the interface region between the rib

and apex, where the material properties change.6. The two exceptions are the BP:BF and the BP:FF models. For these models, the

crack initiates in the rib area only, away from the rib/apex interface.7. For a constant panel mesh, changes in the foam mesh do not substantially affect

the internal and kinetic energy time-history results; however, in all cases, thehourglass energy of the foam is highest for the coarse mesh and lowest for thefine mesh.

8. The magnitude and duration of the contact force time-histories for all cases werenearly identical; however, the contact responses of the coarse foam modelsgenerally exhibited high-frequency oscillations and the contact responses of theFP:FF models contained large spikes that were attributed to instabilities in thecontact algorithm.

9. The baseline mesh discretization is a good choice for future simulations in that itcaptures the structural behavior and damage progression of the panel and foamwithout the computational expense of the fine mesh discretization.

10. The coarse foam mesh should be avoided due to the high levels of hourglassenergy in the model.

11. The input of accurate RCC material properties to the model is extremelyimportant in correctly predicting the initiation and amount of impact damage tothe panel.

References

1. Gehman, H. W., et al, “Columbia Accident Investigation Board,” Report Volume 1,

U. S. Government Printing Office, Washington, DC, August 2003.

2. Anon., “LS-DYNA Keyword User’s Manual Volume I and II – Version 960,”

Livermore Software Technology Company, Livermore, CA, March 2001.

3. Carney, K., Melis, M., Fasanella, E., Lyle, K, and Gabrys, J.: “Material Modeling of

Space Shuttle Leading Edge and External Tank Materials for Use in the Columbia

Accident Investigation.” Proceedings of 8th

International LS-DYNA User’s Conference,

Dearborn, MI, May 2-4, 2004.

4. Melis, M.; Carney, K.; Gabrys, J.; Fasanella, E.; and Lyle, K.: “A Summery of the

Space Shuttle Columbia Tragedy and the Use of LS-DYNA in the Accident Investigation

and Return to Flight Efforts.” Proceedings of 8th

International LS-DYNA User’s

Conference, Dearborn, MI, May 2-4, 2004.

27

5. Gabrys, J.; Schatz, J.; Carney, K.; Melis, M.; Fasanella, E.; and Lyle, K.: “The Use of

LS-DYNA in the Columbia Accident Investigation.” Proceedings of 8th

International LS-

DYNA User’s Conference, Dearborn, MI, May 2-4, 2004.

6. Lyle, K.; Fasanella, E.; Melis, M.; Carney, K.; and Gabrys, J.: “Application of Non-

Deterministic Methods to Assess Modeling Uncertainties for Reinforced Carbon-Carbon

Debis Impacts.” Proceedings of 8th

International LS-DYNA User’s Conference,

Dearborn, MI, May 2-4, 2004.

7. Fasanella, E. L., Lyle, K. H., Gabrys, J., Melis, M., and Carney, K., “Test and Analysis

Correlation of Foam Impact onto Space Shuttle Wing Leading Edge RCC Panel 8,”

Proceedings of 8th

International LS-DYNA User’s Conference, Dearborn, MI, May 2-4,

2004.

REPORT DOCUMENTATION PAGE Form ApprovedOMB No. 0704-0188

2. REPORT TYPE

Technical Memorandum 4. TITLE AND SUBTITLE

The Influence of Mesh Density on the Impact Response of a Shuttle Leading-Edge Panel Finite Element Simulation

5a. CONTRACT NUMBER

6. AUTHOR(S)

Jackson; Karen E.; Fasanella, Edwin L.; Lyle, Karen H.; and Spellman, Regina L.

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

NASA Langley Research Center U.S. Army Research Laboratory Hampton, VA 23681-2199 Vehicle Technology Directorate NASA Langley Research Center Hampton, VA 23681-2199

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)

National Aeronautics and Space AdministrationWashington, DC 20546-0001and U.S. Army Research LaboratoryAdelphi, MD 20783-1145

8. PERFORMING ORGANIZATION REPORT NUMBER

L-19059

10. SPONSOR/MONITOR'S ACRONYM(S)

NASA

13. SUPPLEMENTARY NOTESAn electronic version can be found at http://techreports.larc.nasa.gov/ltrs/ or http://ntrs.nasa.gov

12. DISTRIBUTION/AVAILABILITY STATEMENTUnclassified - UnlimitedSubject Category 05Availability: NASA CASI (301) 621-0390

19a. NAME OF RESPONSIBLE PERSON

STI Help Desk (email: [email protected])

14. ABSTRACT

A study was performed to examine the influence of varying mesh density on an LS-DYNA simulation of a foam projectile impacting the space shuttle leading edge reinforced-carbon-carbon (RCC) Panel 6. Nine cases were executed with all possible combinations of coarse, baseline, and fine meshes of the foam and panel. For each simulation, the same material properties and impact conditions were specified and only the mesh density was varied. In the baseline model, the shell elements representing the RCC panel are 0.2-in. on edge, whereas the foam elements are about 0.5-in. on edge. The element nominal edge-length for the baseline panel was halved to create a fine panel mesh and doubled to create a coarse panel mesh. In addition, the element nominal edge-length of the baseline foam projectile was halved to create a fine foam mesh and doubled to create a coarse foam mesh. The initial impact velocity of the foam was 775 ft/s. Contour plots of resultant panel displacement and effective stress in the foam were compared at five discrete time intervals. Also, selected time-history responses were plotted to determine the influence of mesh density. As a final comparison, the model with a fine panel and fine foam mesh was executed with slightly different RCC material properties.

15. SUBJECT TERMS

Impact simulation; Mesh density; Foam projectile; Space shuttle; LS-DYNA

18. NUMBER OF PAGES

32

19b. TELEPHONE NUMBER (Include area code)

(301) 621-0390

a. REPORT

U

c. THIS PAGE

U

b. ABSTRACT

U

17. LIMITATION OF ABSTRACT

UU

Prescribed by ANSI Std. Z39.18Standard Form 298 (Rev. 8-98)

3. DATES COVERED (From - To)

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

23-376-70-30-07

11. SPONSOR/MONITOR'S REPORT NUMBER(S)

NASA/TM-2004-213501 ARL-TR-3337

16. SECURITY CLASSIFICATION OF:

The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number.PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.

1. REPORT DATE (DD-MM-YYYY)

11 - 200401-

the influence of mesh density on the impact response of a

Documents